Recently, inhibin-A has been shown to be a useful new prenatal marker of Down's syndrome, significantly increasing detection rates. While the placenta is believed to be the major source of inhibin in pregnancy, there are actually very limited data available on specific inhibin dimers in pregnancy. Using a sensitive and specific ELISA we have measured the inhibin-A content of amniotic fluid (AF) to investigate further the biology of inhibin-A in chromosomally normal and abnormal pregnancies. AF from 51 Down's syndrome and 161 chromosomally normal pregnancies between 16 and 19 weeks of gestation were analysed, blinded as to whether the sample was from a Down's syndrome or normal pregnancy. There were no sex differences in inhibin-A content in either the control or Down's syndrome pregnancies. The median (10th-90th percentiles) inhibin-A level in the control pregnancies increased from 339.6 (175.2-649.1) pg/ml at 16 weeks to 592.9 (256.4-1027.3) pg/ml at 19 weeks of gestation. The median (95% confidence interval) inhibin-A in the Down's syndrome pregnancies, expressed as multiples of the median (MoM) to correct for gestation, was 0.77 (0.68-0.89) MoM, significantly lower than the controls (p < 0.001, Mann-Whitney U test). We believe that these data are compatible with more than one source of inhibin-A in pregnancy and suggest that the fetal membranes may be contributing significantly to AF inhibin-A content. Further, our data would suggest that the endocrine function of the placenta and the other inhibin source(s) are differentially regulated.